[Sursound] Ambisonics 2012 - deadline extended
We've had a decent number of abstracts submitted for this, but just in case you hadn't noticed, we extended the deadline to next Monday (19th) as a result of some requests from possible contributors - so, if you didn't manage to get your abstract in in time and thought you'd missed the boat - now is the time to re-start typing! Dave -- These are my own views and may or may not be shared by my employer /*/ /* Dave Malham http://music.york.ac.uk/staff/research/dave-malham/ */ /* Music Research Centre */ /* Department of Music"http://music.york.ac.uk/"; */ /* The University of York Phone 01904 432448*/ /* Heslington Fax 01904 432450*/ /* York YO10 5DD */ /* UK 'Ambisonics - Component Imaging for Audio' */ /*"http://www.york.ac.uk/inst/mustech/3d_audio/"; */ /*/ ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] UHJ decoding and shelf filters
On 15 Dec 2011, at 17:00, sursound-requ...@music.vt.edu wrote: > > Message: 3 > Date: Thu, 15 Dec 2011 14:57:06 + > From: Fons Adriaensen > Subject: Re: [Sursound] UHJ decoding and shelf filters > To: sursound@music.vt.edu > Message-ID: <20111215145706.ga29...@linuxaudio.org> > Content-Type: text/plain; charset=us-ascii > > On Thu, Dec 15, 2011 at 07:19:40PM -, Richard Lee wrote: > >> Fig 3.3 shows them in the form that Ambdec uses; matched gain at LF. >> >> So 'order gain' at LF is 1.0 for both. >> >> At HF, 'order gain' is >> >> 3.7dB for W >> and >> -2.1dB for X & Y > > So what this amounts to is that for normal B-format the difference between > the > HF gains for resp. W and X,Y is 3 dB, and for UHJ it should be 5.8 dB. > > The actual shelf gains (the difference between LF and HF gain > for any component) don't tell the real story. Changing the LF > or HF gain for all components by the same amount doesn't modify > the decoding, it just results in some EQ. > > The Ambdec presets actually normalise for pressure at LF > and for energy at HF. That means that the shapes of the > equivalent shelf filters depend on the number of speakers. > > For example, the square decoder has order gains > +3, 0 at HF, while the octagon first order one > has +6, +3. In both cases the difference between > W and X,Y is 3 dB, this makes the decoder max rE. > > The balance between the HF and LF parts (that is, > the actual slopes of the 'shelf filters') can be > changed at run time using the LF/HF balance control. > > Ambdec doesn't use shelf filters, it uses crossover > filters and two independent matrices. That is still a shelf filter in its composite response! We used such bandsplitting shelf filters right from the earliest experimental Ambisonic decoders in the 1970s as it was much easier to change the cross-over frequency this way. I have always built experimental decoders this way. Its really only when you seek to reduce costs in analogue implementations that shelf filters of the all-pass type become attractive. Geoffrey > So all the > above is assuming that the two matrices are the > same. They usually are for regular layouts, but > not in the case of the 5.1 one. This complicates > matters. > > Looking at the HF order gains for the 5.1 decoder > it seems the ratio is close to 5.8 dB, but this > doesn't mean anything as the matrices are quite > different. > > Ciao, -- next part -- An HTML attachment was scrubbed... URL: <https://mail.music.vt.edu/mailman/private/sursound/attachments/20111216/ac70a025/attachment.html> ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] UHJ decoding and shelf filters
On Fri, Dec 16, 2011 at 09:34:04AM -, Richard Lee wrote:
> Does anyone have Fons' 5.1 matrices handy that they could
> send me? I can't promise anything quick as its 35C in the
> shade and 100% RH in Cooktown and my small brain is boiling.
Here (Parma) it's around 3C and 100% RH (thick mist) in the
evening. I'd prefer Cooktown (provided there's some sea nearby).
This is the complete preset file:
-
# AmbDec configuration
# Written by Makedec-0.8.0 at Sun Sep 4 22:29:44 2011
/description ITU 5.1 decoder
/version 3
/dec/chan_maskb
/dec/freq_bands 2
/dec/speakers 5
/dec/coeff_scale fuma
/opt/input_scale fuma
/opt/nfeff_comp input
/opt/delay_comp off
/opt/level_comp off
/opt/xover_freq600
/opt/xover_ratio 0.0
/speakers/{
add_spkrLS 1.500110.0 0.0system:playback_1
add_spkrLF 1.500 30.0 0.0system:playback_2
add_spkrCE 1.500 0.0 0.0system:playback_3
add_spkrRF 1.500-30.0 0.0system:playback_4
add_spkrRS 1.500 -110.0 0.0system:playback_5
/}
/lfmatrix/{
order_gain 1.0 1.0 1.0 1.0
add_row 0.512590 0.414680 -0.396620
add_row 0.143330 0.220650 0.240850
add_row 0.102380 0.00 0.311540
add_row 0.143330 -0.220650 0.240850
add_row 0.512590 -0.414680 -0.396620
/}
/hfmatrix/{
order_gain 2.05000 1.13000 1.0 1.0
add_row 0.312680 0.252960 -0.241940
add_row 0.172000 0.264780 0.289010
add_row 0.071660 0.00 0.218080
add_row 0.172000 -0.264780 0.289010
add_row 0.312680 -0.252960 -0.241940
/}
/end
-
The matrix columns are W,X,Y. W is multiplied
by the first order_gain, X and Y by the second.
Ciao,
--
FA
Vor uns liegt ein weites Tal, die Sonne scheint - ein Glitzerstrahl.
___
Sursound mailing list
Sursound@music.vt.edu
https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] UHJ decoding and shelf filters
On Fri, Dec 16, 2011 at 3:33 AM, Fons Adriaensen wrote:
> The matrix columns are W,X,Y.
I gather that Ambdec version 3 files (v. 0.5.1) use ACN for the
coefficient ordering, which would make it W, Y, X.
For example the LF coefficients you give for the CE speaker (dir=[1,0,0]) are
> add_row 0.102380 0.00 0.311540
It's in the high-50s (F) and about 40% RH here in Menlo Park.
--
Aaron Heller (hel...@ai.sri.com)
Menlo Park, CA US
On Fri, Dec 16, 2011 at 3:33 AM, Fons Adriaensen wrote:
> On Fri, Dec 16, 2011 at 09:34:04AM -, Richard Lee wrote:
>
>> Does anyone have Fons' 5.1 matrices handy that they could
>> send me? I can't promise anything quick as its 35C in the
>> shade and 100% RH in Cooktown and my small brain is boiling.
>
> Here (Parma) it's around 3C and 100% RH (thick mist) in the
> evening. I'd prefer Cooktown (provided there's some sea nearby).
>
> This is the complete preset file:
>
> -
>
> # AmbDec configuration
> # Written by Makedec-0.8.0 at Sun Sep 4 22:29:44 2011
>
> /description ITU 5.1 decoder
>
> /version 3
>
> /dec/chan_mask b
> /dec/freq_bands 2
> /dec/speakers 5
> /dec/coeff_scale fuma
>
> /opt/input_scale fuma
> /opt/nfeff_comp input
> /opt/delay_comp off
> /opt/level_comp off
> /opt/xover_freq 600
> /opt/xover_ratio 0.0
>
> /speakers/{
> add_spkr LS 1.500 110.0 0.0 system:playback_1
> add_spkr LF 1.500 30.0 0.0 system:playback_2
> add_spkr CE 1.500 0.0 0.0 system:playback_3
> add_spkr RF 1.500 -30.0 0.0 system:playback_4
> add_spkr RS 1.500 -110.0 0.0 system:playback_5
> /}
>
> /lfmatrix/{
> order_gain 1.0 1.0 1.0 1.0
> add_row 0.512590 0.414680 -0.396620
> add_row 0.143330 0.220650 0.240850
> add_row 0.102380 0.00 0.311540
> add_row 0.143330 -0.220650 0.240850
> add_row 0.512590 -0.414680 -0.396620
> /}
>
> /hfmatrix/{
> order_gain 2.05000 1.13000 1.0 1.0
> add_row 0.312680 0.252960 -0.241940
> add_row 0.172000 0.264780 0.289010
> add_row 0.071660 0.00 0.218080
> add_row 0.172000 -0.264780 0.289010
> add_row 0.312680 -0.252960 -0.241940
> /}
>
>
> /end
>
>
> -
>
> The matrix columns are W,X,Y. W is multiplied
> by the first order_gain, X and Y by the second.
>
> Ciao,
>
> --
> FA
>
> Vor uns liegt ein weites Tal, die Sonne scheint - ein Glitzerstrahl.
>
> ___
> Sursound mailing list
> Sursound@music.vt.edu
> https://mail.music.vt.edu/mailman/listinfo/sursound
___
Sursound mailing list
Sursound@music.vt.edu
https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] UHJ decoding and shelf filters
On 16/12/11 17:48, Aaron Heller wrote: It's in the high-50s (F) and about 40% RH here in Menlo Park It zero degrees here at Hest Bank on the edge of Morecambe Bay in the north of England and the water in the air has turned solid. I will need some time to digest all this information. It might be moot anyway if I cannot fix my other current problem with the ac3/a52 encoded speaker feeds. Either my encoder (aften) is severely broken or whatever I try to play them back on wants to remix all the channels in a way the completely destroys any ambisonic localisation. I will have to wait till the house is empty before I can start playing test tones through a real DVD player :-). Merry Christmas Roger ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] UHJ decoding and shelf filters
On Fri, Dec 16, 2011 at 09:48:17AM -0800, Aaron Heller wrote: > On Fri, Dec 16, 2011 at 3:33 AM, Fons Adriaensen wrote: > > The matrix columns are W,X,Y. > > I gather that Ambdec version 3 files (v. 0.5.1) use ACN for the > coefficient ordering, which would make it W, Y, X. Absolutely right - version 3 uses ACN. Thanks for the correction ! Ciao, -- FA Vor uns liegt ein weites Tal, die Sonne scheint - ein Glitzerstrahl. ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] UHJ decoding and shelf filters
Is BLaH4 available anywhere apart from the AES site? I am not a member, only curious. Gerard Lardner On 16/12/2011 09:34, Richard Lee wrote: > BLaH4, "Design of Ambisonic Decoders for Irregular Arrays of > Loudspeakers by Non- Linear Optimization" Heller et al, preprint 8243 > AES San Francisco nov10, discusses the various trade-offs and presents > our favoured method. ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
